diff options
Diffstat (limited to 'drivers/gpu/drm/amd/amdkfd/kfd_chardev.c')
-rw-r--r-- | drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 90 |
1 files changed, 57 insertions, 33 deletions
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index 2b3d8bc8f0aa..6d291aa6386b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -327,6 +327,12 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p, goto err_bind_process; } + if (!pdd->doorbell_index && + kfd_alloc_process_doorbells(dev, &pdd->doorbell_index) < 0) { + err = -ENOMEM; + goto err_alloc_doorbells; + } + /* Starting with GFX11, wptr BOs must be mapped to GART for MES to determine work * on unmapped queues for usermode queue oversubscription (no aggregated doorbell) */ @@ -404,6 +410,7 @@ err_create_queue: if (wptr_bo) amdgpu_amdkfd_free_gtt_mem(dev->adev, wptr_bo); err_wptr_map_gart: +err_alloc_doorbells: err_bind_process: err_pdd: mutex_unlock(&p->mutex); @@ -869,14 +876,11 @@ static int kfd_ioctl_wait_events(struct file *filp, struct kfd_process *p, void *data) { struct kfd_ioctl_wait_events_args *args = data; - int err; - err = kfd_wait_on_events(p, args->num_events, + return kfd_wait_on_events(p, args->num_events, (void __user *)args->events_ptr, (args->wait_for_all != 0), - args->timeout, &args->wait_result); - - return err; + &args->timeout, &args->wait_result); } static int kfd_ioctl_set_scratch_backing_va(struct file *filep, struct kfd_process *p, void *data) @@ -1092,6 +1096,10 @@ static int kfd_ioctl_alloc_memory_of_gpu(struct file *filep, goto err_unlock; } offset = kfd_get_process_doorbells(pdd); + if (!offset) { + err = -ENOMEM; + goto err_unlock; + } } else if (flags & KFD_IOC_ALLOC_MEM_FLAGS_MMIO_REMAP) { if (args->size != PAGE_SIZE) { err = -EINVAL; @@ -1576,6 +1584,8 @@ static int kfd_ioctl_smi_events(struct file *filep, return kfd_smi_event_open(pdd->dev, &args->anon_fd); } +#if IS_ENABLED(CONFIG_HSA_AMD_SVM) + static int kfd_ioctl_set_xnack_mode(struct file *filep, struct kfd_process *p, void *data) { @@ -1586,22 +1596,29 @@ static int kfd_ioctl_set_xnack_mode(struct file *filep, if (args->xnack_enabled >= 0) { if (!list_empty(&p->pqm.queues)) { pr_debug("Process has user queues running\n"); - mutex_unlock(&p->mutex); - return -EBUSY; + r = -EBUSY; + goto out_unlock; } - if (args->xnack_enabled && !kfd_process_xnack_mode(p, true)) + + if (p->xnack_enabled == args->xnack_enabled) + goto out_unlock; + + if (args->xnack_enabled && !kfd_process_xnack_mode(p, true)) { r = -EPERM; - else - p->xnack_enabled = args->xnack_enabled; + goto out_unlock; + } + + r = svm_range_switch_xnack_reserve_mem(p, args->xnack_enabled); } else { args->xnack_enabled = p->xnack_enabled; } + +out_unlock: mutex_unlock(&p->mutex); return r; } -#if IS_ENABLED(CONFIG_HSA_AMD_SVM) static int kfd_ioctl_svm(struct file *filep, struct kfd_process *p, void *data) { struct kfd_ioctl_svm_args *args = data; @@ -1621,6 +1638,11 @@ static int kfd_ioctl_svm(struct file *filep, struct kfd_process *p, void *data) return r; } #else +static int kfd_ioctl_set_xnack_mode(struct file *filep, + struct kfd_process *p, void *data) +{ + return -EPERM; +} static int kfd_ioctl_svm(struct file *filep, struct kfd_process *p, void *data) { return -EPERM; @@ -1928,7 +1950,7 @@ static int criu_checkpoint(struct file *filep, { int ret; uint32_t num_devices, num_bos, num_objects; - uint64_t priv_size, priv_offset = 0; + uint64_t priv_size, priv_offset = 0, bo_priv_offset; if (!args->devices || !args->bos || !args->priv_data) return -EINVAL; @@ -1972,38 +1994,34 @@ static int criu_checkpoint(struct file *filep, if (ret) goto exit_unlock; - ret = criu_checkpoint_bos(p, num_bos, (uint8_t __user *)args->bos, - (uint8_t __user *)args->priv_data, &priv_offset); - if (ret) - goto exit_unlock; + /* Leave room for BOs in the private data. They need to be restored + * before events, but we checkpoint them last to simplify the error + * handling. + */ + bo_priv_offset = priv_offset; + priv_offset += num_bos * sizeof(struct kfd_criu_bo_priv_data); if (num_objects) { ret = kfd_criu_checkpoint_queues(p, (uint8_t __user *)args->priv_data, &priv_offset); if (ret) - goto close_bo_fds; + goto exit_unlock; ret = kfd_criu_checkpoint_events(p, (uint8_t __user *)args->priv_data, &priv_offset); if (ret) - goto close_bo_fds; + goto exit_unlock; ret = kfd_criu_checkpoint_svm(p, (uint8_t __user *)args->priv_data, &priv_offset); if (ret) - goto close_bo_fds; + goto exit_unlock; } -close_bo_fds: - if (ret) { - /* If IOCTL returns err, user assumes all FDs opened in criu_dump_bos are closed */ - uint32_t i; - struct kfd_criu_bo_bucket *bo_buckets = (struct kfd_criu_bo_bucket *) args->bos; - - for (i = 0; i < num_bos; i++) { - if (bo_buckets[i].alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_VRAM) - close_fd(bo_buckets[i].dmabuf_fd); - } - } + /* This must be the last thing in this function that can fail. + * Otherwise we leak dmabuf file descriptors. + */ + ret = criu_checkpoint_bos(p, num_bos, (uint8_t __user *)args->bos, + (uint8_t __user *)args->priv_data, &bo_priv_offset); exit_unlock: mutex_unlock(&p->mutex); @@ -2145,6 +2163,12 @@ static int criu_restore_devices(struct kfd_process *p, ret = PTR_ERR(pdd); goto exit; } + + if (!pdd->doorbell_index && + kfd_alloc_process_doorbells(pdd->dev, &pdd->doorbell_index) < 0) { + ret = -ENOMEM; + goto exit; + } } /* @@ -2173,6 +2197,8 @@ static int criu_restore_memory_of_gpu(struct kfd_process_device *pdd, return -EINVAL; offset = kfd_get_process_doorbells(pdd); + if (!offset) + return -ENOMEM; } else if (bo_bucket->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_MMIO_REMAP) { /* MMIO BOs need remapped bus address */ if (bo_bucket->size != PAGE_SIZE) { @@ -2847,7 +2873,6 @@ static int kfd_mmio_mmap(struct kfd_dev *dev, struct kfd_process *process, struct vm_area_struct *vma) { phys_addr_t address; - int ret; if (vma->vm_end - vma->vm_start != PAGE_SIZE) return -EINVAL; @@ -2867,12 +2892,11 @@ static int kfd_mmio_mmap(struct kfd_dev *dev, struct kfd_process *process, process->pasid, (unsigned long long) vma->vm_start, address, vma->vm_flags, PAGE_SIZE); - ret = io_remap_pfn_range(vma, + return io_remap_pfn_range(vma, vma->vm_start, address >> PAGE_SHIFT, PAGE_SIZE, vma->vm_page_prot); - return ret; } |